Hi;
We have encountered a performance issue with buffermapping LWLock in PG14.3(on CentOS7 ~ modern machine) .
Sometimes we see that the buffermaping wait event increases and after a while all sessions are waiting for this lock (as far as we understand there is a threshold value about buffermapping lwlock, all sessions start to wait when it passes).
How did those who faced this problem go about? both on the database side and the application side?
regards...